-
2013 International Monetary Fund
IMF Country Report No. 13/155
GREECE SELECTED ISSUES
This Selected Issues paper for Greece was prepared by a staff
team of the International Monetary Fund as background documentation
for the periodic consultation with the member country. It is based
on the information available at the time it was completed on May
21, 2013. The views expressed in this document are those of the
staff team and do not necessarily reflect the views of the
government of Greece or the Executive Board of the IMF.
The policy of publication of staff reports and other documents
allows for the deletion of market-sensitive information.
Copies of this report are available to the public from
International Monetary Fund Publication Services 700 19th
Street, N.W. Washington, D.C. 20431
Telephone: (202) 623-7430 Telefax: (202) 623-7201 E-mail:
[email protected] Internet: http://www.imf.org
International Monetary Fund Washington, D.C.
June 2013
-
GREECE SELECTED ISSUES Approved By The European Department
Prepared by Stephanie Eble, Wojciech Maliszewski, Maral Shamloo
(all EUR), and Iva Petrova (FAD).
CONTENTS
RESTORING GROWTH
_______________________________________________________________ 3A.
What Explains Greeces Adjustment Pattern?
___________________________________________ 3B. Can Structural
Reforms Reinvigorate and Sustain Growth?
_____________________________ 7C. Simulation Results
_____________________________________________________________________
11D. Conclusions
___________________________________________________________________________
13 BOX 1. GIMF Model
_____________________________________________________________________________
9 FIGURE 1. Simulation Results, 201320
___________________________________________________________
14REFERENCES
________________________________________________________________________
16REVENUE ADMINISTRATION AND FISCAL CONSOLIDATION
____________________ 18A. Why Improving Revenue Administration is
Critical for Greece ________________________ 18B. Revenue Losses in
Greece and the Scope for Higher Collection ______________________
19C. What Explains the Large Amount of Tax Evasion in Greece?
__________________________ 23D. The Way Forward
_____________________________________________________________________
28E. Change: How and How Quickly Can It Happen?
_______________________________________ 32F. Conclusions
____________________________________________________________________________
36 FIGURES 1. Tax EvasionsA Decision Tree
________________________________________________________ 252.
Revenue Administration Reform Strategy
_____________________________________________ 30
May 21, 2013
-
GREECE
2 INTERNATIONAL MONETARY FUND
TABLES 1. Where is Greece Losing All Its Tax Revenue?
__________________________________________ 212. Costs of
Alternative Tax Payment Schemes: Bank Loan, Deferral Scheme or
Payment Following Audits
_________________________________________________________________________
253. Irrationality of Paying Taxes
___________________________________________________________ 274.
Criteria for Tax Administration Reform
________________________________________________ 295. Country
Experiences with Institutional Reforms
_______________________________________ 33REFERENCES
________________________________________________________________________
38HOW CAN THE FINANCIAL SECTOR SUPPORT RECOVERY?
______________________ 40A. Introduction
___________________________________________________________________________
40B. Creditless Recoveries: Is there a Contradiction?
_______________________________________ 40C. Policies
________________________________________________________________________________
47D. Conclusions
___________________________________________________________________________
55 BOXES 1. Credit Projections in the Macroframework
____________________________________________ 432. Estimating the
Likelihood of a Creditless Recovery for Greece
________________________ 443. Real Estate Market Reforms
___________________________________________________________ 514.
World Bank Principles B3 and B5.2 for Out-of-Court Restructuring
Mechanisms _____ 535. Latvia: Out of Court Company Debt
Restructuring Principles _________________________ 54 FIGURES1.
Financial Sector Projections, 200116
_________________________________________________ 562. Credit and
GDP Evolution in Latvia, Iceland, and Ireland, 20102012
_________________ 573. Non-Bank Financing and Sectoral
Reallocation, 200513 _____________________________ 58REFERENCES
________________________________________________________________________
61
-
GREECE
INTERNATIONAL MONETARY FUND 3
RESTORING GROWTH1 Rigidities in Greeces product and labor
markets have increased the cost of adjustment to large pre-crisis
economic imbalances. Simulations from a calibrated model of the
Greek economy confirm that reforms to these markets can play a
significant role in stemming output losses and supporting the
recovery. In particular, with important labor market reforms having
been implemented recently, steadfast implementation of product
market reforms is key to reinvigorating growth.
A. What Explains Greeces Adjustment Pattern?
1. Greece has made headway in restoring competitiveness via
internal devaluation, but adjustment costs have been high. The
improvement in the trade balance by 9 percentage points since 2009
was achieved by contracting output by 22 percenta much worse
trade-off than in other economies. Despite initially encouraging
performance, export growth has done little to cushion the impact of
domestic demand collapse on the economy. By contrast, other crisis
countries (Spain, Portugal, Ireland) have had relatively more
success in stimulating exports and avoiding a deep output collapse.
The difference can be explained in part by Greeces larger initial
imbalances (e.g., the CGER-estimated REER misalignment was higher
for Greece than for Spain and Portugal), and hence by the need for
a greater reallocation of resources. Such reallocation is
inevitably costly in the short-run. Even so, the loss of output has
been proportionately higher in Greece.
2. To account for the high adjustment costs, we conduct an
empirical analysis of growth during large adjustment episodes. Our
econometric analysis follows a large literature on growth
performance in current account adjustment episodes (e.g.,
Milesi-Ferretti and Razin, 1998), with two important distinctions.
First, we choose a sample based on large domestic demand
contractions rather than large current account adjustments. This
focuses the analysis on adjustments to negative shocks and
eliminates externally-driven current account improvements. Second,
we add structural indicators to test for the potential effect of
labor and product market rigidities on growth during the adjustment
episodes. Adjustment episodes are defined by two criteria: (i)
negative domestic demand growth; and (ii) a contraction in the
average domestic-absorption-to-GDP ratio by more than 3 percentage
points between two consecutive three-year periods. To avoid
excess
1 Prepared by Wojciech Maliszewski.
-
GREECE
4 INTERNATIONAL MONETARY FUND
heterogeneity, the sample is restricted to OECD and emerging
market economies. In this exercise, the variables used to explain
cumulative growth are as follows:
Institutional indicators. Product and labor market rigidities
raise adjustment costs by hindering the reallocation of resources.
In our empirical analysis, we include labor and product market
regulations at the start of an adjustment episode as proxied,
respectively, by the Labor market regulations index and the Freedom
to trade internationally index, both from Economic Freedom of the
World Annual Report (Gwartney et al., 2012). Note that, while
better indicators of product market regulations are available
(e.g., OECD indices of product market regulations), their coverage
of countries and of time samples is narrower. We use country
rankings in both casesa higher value indicates more regulated
product and labor markets.
Exchange rate regime. Fixed regimes limit or eliminate the
nominal exchange rate adjustment channel, potentially increasing
costs. We use a dummy for fixed exchange rate regimes based on the
IMFs Annual Report on Exchange Arrangements and Exchange
Restrictions (AREAER, 2012). Pegged regimes are defined as
belonging to three AREAER categories: no separate legal tender,
currency board, and conventional pegs. Members of currency unions
are also classified as operating under fixed exchange rate.
Exchange rate misalignment. The extent of initial imbalances
dictates the size and possibly the speed of the subsequent
adjustment. We use the IMFs CGER measure of exchange rate
misalignment, and specifically the equilibrium real exchange rate
method (Lee et al, 2008).
Financial conditions are potentially an important determinant of
adjustment costs, determining both the degree of consumption
smoothing and support for investments. Give the potential
endogeneity of indicators of financial conditions, and given the
focus of the analysis on large domestic demand contractions, we use
a dummy indicating banking crisis episodes during or up to two
years before the beginning of the adjustment episode (Laeven and
Valencia, 2012) as a proxy for the tightening of financing
conditions.
External shocks. External conditions are approximated by growth
in partner countries (weighted by trade shares). This series, as
well as cumulative GDP growth and national accounts data used to
select the episodes, are from the IMFs WEO database.
-
GREECE
INTERNATIONAL MONETARY FUND 5
3. The econometric results show that adjustment costs are shaped
both by initial conditions and macroeconomic shocks and by the
degree of product and labor market flexibility (Table). Although
based on a small sample of adjustment episodes, the results
indicate that the degree of overvaluation, external demand growth,
and the tightening of financial conditions in the aftermath of
banking crises are strongly and significantly linked to adjustment
costs. Product and labor market regulations also have a negative
and significant impact on growth, and the estimates are fairly
robust (as indicated by alternative estimation methods). Pegged
regimes also seem to be negatively correlated with growth during
the adjustment episode, as indicated by the high and negative
coefficient of the peg dummy. To further test whether adjustments
under fixed exchange rate regimespotentially more relevant to draw
lessons for Greecediffer from full sample results, the model is
re-estimated based on a sample restricted to those episodes under
fixed exchange rate regimes. The estimates are less
precisesuggesting a wide range of outcomes under fixed exchange
rate regimes in the samplebut values and signs of key coefficients
remain close to the full sample estimation.
4. Greeces low score in both product and labor market
flexibility indicators suggests that structural rigidities have
played an important role in making the adjustment more costly.
While labor market reforms have advanced since the beginning of the
crisis, there has been relatively little progress in addressing
structural rigidities in product markets. Several factors played a
role in restraining product and labor market flexibility:
(1) (2) (3)
Indicator OLS Robust Regression 2/ OLS
Freedom to trade internationally (rank) -0.086* -0.093*
-0.068(0.045) (0.050) (0.064)
Labor market regulations (rank) -0.085* -0.099* -0.083(0.049)
(0.053) (0.077)
Exchange rate peg (dummy) -2.817 -3.252* ...(1.700) (1.852)
...
Exchange rate overvaluation (percent) -0.130*** -0.132***
-0.087(0.044) (0.048) (0.063)
External demand (percent change) 0.373** 0.339* 0.643**(0.178)
(0.194) (0.264)
Banking sector crisis (percent) -6.581*** -6.847***
-3.809(1.774) (1.933) (3.137)
Constant 3.090 4.343 -0.935(2.840) (3.094) (3.729)
Observations 42 42 27R-squared 0.580 0.558 0.460
Source: IMF staff estimates.
2/ Robust regression reestimates the baseline specification
using robust regression, which downweights observations with larger
absolute residuals using iterative weighted least squares
(Andersen, 2008).
Explaining Cumulative Growth During Episodes of Large Domestic
Demand Contractions 1/
1/ Standard errors in parentheses; *** p
-
GREECE
6 INTERNATIONAL MONETARY FUND
Greece entered the crisis with an overburden of regulation. The
key sources include: (i) important assets have been ownedor their
use restrictedby the state, and they remain underutilized (e.g.,
because of land zoning restrictions) or subject to monopolies
(network utilities); (ii) regulations have been excessive, with
permitting, licensing and export-import requirements well inferior
to EU best practice; and (iii) even beyond explicit requirements,
outsiders have been discouraged from entry into new markets by
implicit barriers, including excessive bureaucracy, corruption, and
opaque tax and labor regulations. These barriers limited
competition (including from FDI), keeping prices and margins well
above the EU average level, and preserving an economic model based
on small and inefficient enterprises.
Similarly, Greeces labor market regulations were rigid and
tended to protect insiders. The labor market has traditionally
suffered from a closed and inflexible system of collective
bargaining, very high firing costs (severance payments and
redundancy notification periods), a high national minimum wage
relative to competitors, and high non-wage labor costs. The reforms
under the program have significantly reduced the rigidities, which
facilitated wage adjustments and contributed to the reduction in
unit labor costs (ULCs).
-
GREECE
INTERNATIONAL MONETARY FUND 7
5. Would the elimination of existing distortions limit output
losses during the recession and contribute to the recovery? While
econometric estimates give a broad-brush picture of the correlation
between structural rigidities and growth performance during the
adjustment, they are not sufficient to analyze the effects of
removing the distortions: the data sample is limited and proxies
for potential explanatory factors are imperfect. Given the limited
data, the model does not directly capture the response of growth to
reforms reducing structural rigidities (correlating it instead only
with a level). Neither does it consider the sequencing of policies
(e.g., labor and product market reforms) that could potentially
affect adjustment costs in different ways. To explore these
questions, we use a calibrated model of the Greek economy in the
next section.
B. Can Structural Reforms Reinvigorate and Sustain Growth?
6. What economic effects can realistically be expected from the
implementation of structural reforms? While the initial program for
Greece relied heavily on productivity gains (driven by structural
reforms) to deliver a quick rebound in growth, the strategy has
been subsequently revised to allow for deeper wage and price
adjustment, more fully taking into account the authorities
implementation capacity and the transmission channels of reforms.
This section addresses two questions: what can structural reforms
deliver in terms of growth and improvement in competitiveness, and
how fast? We review how structural reforms affect economic
performance in theory and previous empirical studies, before
turning to Greece-specific simulations.
Theoretical channels and results from previous empirical
studies
7. Theoretical results point to benefits from structural
reforms, but indicate that they may not materialize
immediately:
The main theoretical channel through which structural reforms
affect economic performance is through a reduction in rents
(Blanchard and Giavazzi, 2001). Removing barriers to entrywhich are
defined broadly, including excess bureaucracy and other
deficiencies in business environmentenhances competition and
reduces rents. This brings prices more in line with marginal costs,
encouraging managers and workers to operate more efficiently.
The process of improving efficiency of the economy is not
frictionless. It involves improving allocative efficiency (moving
resources to more productive uses) and productive efficiency
(organizing work more effectively, trimming fat and reducing
slack). In the process, low- productivity firms exit, releasing
resources to be absorbed by more productive firms (who are also
better able to compete abroad). Transition costs can be high,
particularly when market imperfections in the credit market
constrain investments of productive firms, or uncertainty about
economic prospects weigh on investments directly.
-
GREECE
8 INTERNATIONAL MONETARY FUND
Different forces play a role in increasing long-term
productivity growth. The process of improving allocative and
productive efficiency ultimately comes to a stop when the economy
reaches the production frontier, and innovation and the
introduction of new goods and processes drive growth
thereafter.
8. Empirical results are in line with the theory, indicating
that the effects of structural reforms build up gradually:
In the long run, product and labor market reforms can have
positive effects on growth, employment, and productivity (e.g.,
Bouis and Duval, 2011; Barnes et al; 2011; OECD, 2012; Hobza and
Mourre, 2010);
In the short run, the impact is smaller because of adjustment
costs, especially in the case of labor market reforms (Cacciatore
et al, 2012), and particularly when these are undertaken during
recessions (Bouis et al, 2012);
There are complementarities between product and labor market
reforms: a broad reform package would be more beneficial than
individual reforms as the former could help lower transitional
costs (Cacciatore et al, 2012). They would also improve the
distributional consequences that might otherwise arise if wage
declines are not matched with proportional price declines.
9. Greece-specific empirical results point to potentially
sizeable positive effects of structural reforms on GDP and
productivity. In particular, using a calibrated model, Zonzilos
(2010) finds that product and labor market reforms could boost GDP
level by about 10 percent in the steady state, and also help
restore price competitiveness. This is in line with previous
empirical studies.
Model setup and simulation
10. The impact of product and labor market reforms in Greece is
simulated using the IMFs Global Integrated Monetary and Fiscal
model (GIMF). This DSGE model (see Box 1) can help illuminate
transmission channels and analyze the sequencing of reforms. Given
the presence of monopolistic competition in both firms and labor
markets, the GIMF is well suited to analyze the effects of
structural reforms reducing price and wage markups (because, as
discussed above, structural reforms are usually framed in terms of
making the markets more competitive, for example, through reducing
entry barriers). Our simulation framework is similar to Lusinyan
and Muir (2012).
-
GREECE
INTERNATIONAL MONETARY FUND 9
11. The standard calibration of GIMF is augmented with
Greek-specific information. To reflect the negative impact of the
crisis on the financial system, it is assumed that liquidity
constrained households make up 50 percent of all households
(compared to 25 percent for the euro area). The markups are
calibrated to be consistent with empirical studies and previous
results in the literature: the non-tradable sector price markup is
60 percent versus 35 percent for the rest of the euro area; and 20
percentagainst 15 percent in the euro areafor the tradable sector.
For the wage markup, we follow the assumption in Forni et al.
(2010) and use the same values as for the non-tradable sector price
markup. Steady state ratios are calibrated to 2011 values, with
some modifications. In particular, the persistently high current
account deficit is assumed to be eliminated in the steady state
through higher exports and the debt-to-GDP ratio is assumed to
decline to the euro area level in the steady state.
Box 1. GIMF Model GIMF is a multi-country Dynamic Stochastic
General Equilibrium (DSGE) model with optimizing behavior by
households and firms, and intertemporal stock-flow accounting
(Kumhof et al., 2010):
Frictions in the form of sticky prices and wages, real
adjustment costs, liquidity-constrained households, and
finite-planning horizons of households imply an important role for
monetary and fiscal policy in economic stabilization (the
assumption of finite horizons separates GIMF from standard monetary
DSGE models and allows it to have well-defined steady states where
countries can be long-run debtors or creditors).
The non-Ricardian features of the model provide non-neutrality
in both spending-based and revenue-based fiscal measures, which
makes the model particularly suitable to analyze fiscal policy
questions (fiscal policy can affect the level of economic activity
in the short run).
Government debt is only held domestically, as nominal,
non-contingent, one-period bonds denominated in domestic currency.
The only assets traded internationally are nominal, non-contingent,
one-period bonds denominated in U.S. dollars that can be issued by
the U.S. government and by private agents in any region.
Firms employ capital and labor to produce tradable and
nontradable intermediate goods. They are owned domestically (equity
is not traded in domestic financial markets; instead, households
receive lump-sum dividend payments).
A financial sector a la Bernanke, Gertler, and Gilchrist (1999)
incorporates a procyclical financial accelerator, with the cost of
external finance of firms rising with their indebtedness.
The version of GIMF used in this paper comprises 3 regions:
Greece, the euro area, and the rest of the world.
12. Several simulations are analyzed. In a baseline simulation,
there are no structural reforms. In a second simulation, only labor
market reforms only are analyzed. A third simulation adds to the
second simulation by analyzing product market reforms as well,
albeit with hesitant and delayed implementation of these reforms. A
final simulation looks at the effect of credible, upfront
implementation of both labor and product market reforms.
-
GREECE
10 INTERNATIONAL MONETARY FUND
13. In line with theoretical results and previous studies, the
reforms are modeled as a reduction in rents. We assume that
increased competition in the product and labor markets brings
margins down. The key assumption here is the extent to which
specific reforms could translate into markups. While a one-to-one
mapping of specific reforms to reduction in rents is difficult to
achieve, cross-country econometric results point to significant
effects of reforms on markups. For instance, Griffiths and Harrison
(2004) find that reductions in costs of doing business, regulatory
trade barriers, labor market regulations, and government
bureaucracy significantly reduce price markups in a sample of OECD
countries. E.g., the 1995-2000 reforms in Germany to reduce costs
of starting a business are estimated to have reduced the markup by
2 percentage points. The effects of wage markups are more difficult
to establish empirically, but Stewart (1990) for instance presents
evidence that markups are higher in unionized establishments facing
less competition. Theoretical results also suggest that reforms to
reduce the difficulty of hiring and firing, a lower wage
replacement ratio, and increasing competition in product markets
contribute to lower wage markups.
14. We assume that the reforms close roughly half the gap
between the Greece and the rest of the euro area over a five-year
period. The reforms to labor markets already underway and to
product markets currently in the pipeline are similar to the set of
reforms affecting the markups identified in the empirical
literature. The advances made to date in labor market reforms give
support to this assumption on timing and their implementation path
is assumed to be known with certainty by economic agents. But,
given slow progress on product market reforms, the simulations
distinguish between product market reforms that are not entirely
credible (which limits the upfront benefits from the reforms) from
those that are implemented fully and upfront. In the former,
economic agents regard already implemented changes as permanent,
but view further reforms with uncertainty in the near term.
15. We also assume that structural reforms reduce wage and price
rigidities. While increasing competition ultimately lowers prices
and wages, the speed of change is affected by nominal rigidities.
These are a function of the degree of competition itself, but also
of institutional regulations, such as minimum wages or price
regulations. Before the crisis, price rigidities in Greece appeared
much higher than in the rest of the euro area, while wage
rigidities were in line with other countries. However, with
typically more constraints to deflation and cuts in nominal wages
compared to just slowdowns in price and wage dynamics, the
inability of real wages to adjust downward is more constraining in
severe recessions. We assume that wage and price rigidities are
reduced by about 30 percent as part of the reform process..
-
GREECE
INTERNATIONAL MONETARY FUND 11
16. Importantly, to mimic the recession dynamics observed in
Greece, we add shocks to the model. The fiscal adjustmentmodeled to
match the dynamics of the cyclically-adjusted primary balance in
the adjustmentprimes the downturn. However, it does not generate
sufficient persistence to model Greeces recession. This suggests
that key explanators for Greeces growth performance during the
crisis go well beyond fiscal multipliers to include the range of
severe confidence shocks hitting the economy in 201112, affecting
investments and consumption both directly and through a liquidity
squeeze. To align model-simulated growth path with actual data, we
thus add shocks to investment, consumption, and financing.
C. Simulation Results
17. Labor market reforms bring modest benefits in terms of
output, but significantly reduce current account deficits (Figure
1). Without greater competition in the product market, the
increased flexibility in the labor market has only marginal effects
on prices, which drop by about of a percentage point after 5 years
and by 2 percent by 2030. Output increases are also modest (about
half a percent after 5 years and 3 percent by 2030). Employment
increases though, as lower wage markups induce a switch from
capital to labor in inputs to production. There is a strong
reduction in current account deficits, driven by both reductions in
imports and a modest increase in exports (as initial real wage
declines limit consumption and investment, while supporting demand
for labor and domestic supply).
-
GREECE
12 INTERNATIONAL MONETARY FUND
18. Product market reforms reduce prices and boost output. Lower
barriers to entry encourage competition, reducing prices by 1
percent in the medium term and by over 6 percent by 2030. The
reduction in prices and stronger demand for labor increase real
incomes, boosting consumption. Investments increase in anticipation
of the future stronger demand from the external sector and
consumption. Output growth is strong, averaging about half a
percentage point more than in the baseline without reforms, with
the cumulative effects of over 3 and 6 percentage points after five
years and by 2030, respectively. Higher investments and consumption
would initially lead to a deterioration of the current account,
reversed when the new production capacity comes on stream. The
latter would increase exports: initial gains would be modestgiven
capacity constraintsbut would rise in the medium to long term,
helped by strong competitiveness gains.
19. The results of the two reforms are additive, producing
significantly higher output and lower prices, while improving
external current account with a lag. Output growth is as much as
percentage point per year higher than in the baseline without
reforms, with the cumulative impact of over 10 percent by 2030.
Prices drop by over 6 percent cumulatively during this period. As
noted, the current account deteriorates initially, reflecting
largely the effects of product market reforms on investment and
consumption, only to improve when the new export capacityencouraged
by improved competitivenesscomes on stream.
20. The simulated effects of reforms are in line with
developments in the Greek economy. While identifying the effects of
reforms is made difficult by the number of adverse shocks affecting
the economy and the dynamics of the recession occurring
simultaneously, economic developments following the implementation
of labor market reforms are in line with model predictions: it led
to large reductions in wages combined with the improvement in the
current account through import compression, without markedly
stemming output declines. Given uneven and delayed implementation
of product market reforms, it is not unexpected that output gains
are not visible.
21. The results are also consistent with long-term projected
growth under the program. While the effect of reforms on growth
ultimately taper off, they are still significant well beyond 2020.
The cumulative effect of combined reforms on output reaches about
10 percent after 2030. The gradual accumulation of gains from
reforms supports the projected TFP growth under the program, which
is consistent with previous empirical results indicating relatively
long payoff periods from structural reforms (Barnes et al., 2011).
The assumed medium-term productivity growth in Greece is also
consistent with results reported in other countries undertaking
structural reforms (e.g., percent per year in Germany and 1 percent
in Netherlands). The still high productivity gap to the core of
Europe indicates that the potential for catch-up growth is high,
and gains from reforms could even exceed estimates from the
model.
-
GREECE
INTERNATIONAL MONETARY FUND 13
22. Growth could be frontloaded with more credible reform
implementation. Alternative simulations in which economic agents
fully anticipate future reforms boost growth though expectations of
future recovery. The effect is strong: growth in the first year of
reform implementation (assuming that both labor and product market
reforms are implemented simultaneously) would be 1 percentage point
higher compared to the non-credible implementation of reforms in
our baseline simulations (long-term effects would be close in both
scenarios).
D. Conclusions
23. Greeces high adjustment costs can be explained in part by
structural rigidities. While a number of factors have played a role
in accounting for Greeces growth performance in the past few years,
pervasive structural rigidities have raised the cost of adjustment.
Their effect has been direct and indirect: they likely contributed
to the significant imbalances before the crisis, which made the
adjustment painful given the scale of imbalances that needed to be
corrected. They also likely created disincentives for an efficient
reallocation of resources (by preserving rents) and increasing
price and wage rigidities (delaying the necessary nominal
adjustment).
24. The impact of structural reforms on GDP can be sizeable. We
confirm earlier findings from the literature using the IMFs Global
Integrated Monetary and Fiscal model (GIMF) showing that policies
that would close roughly half the gap in product and labor markets
with the rest of the euro area could raise real GDP by about 4
percent after 5 years and by 10 percent in the long run. This is a
significant boost to the economy, especially beyond the projected
cyclical upturn when the labor force begins to shrink. It is also
in line with previous econometric studies. It is smaller, however,
than gains expected at the initiation of the program in 2010,
reflecting more realistic assumptions about both the pace of
implementation and the transmission channels. With reforms
currently in place, staffs latest projections are in line with
model-based simulations.
25. A decisive implementation of product market reforms would
have a measurable impact on output dynamics, inflation, and
competitiveness. The early implementation of reforms in the labor
market led to sharp reductions in wages, stemmed employment losses,
and reduced current account. But a broader set of product market
reforms is needed to meaningfully reinvigorate growth. Such reforms
would have a meaningful impact on the speed of recovery (as
credible product market reforms encourage investments in
anticipation of future gains and exports boosted by the new
capacity and improvements in competitiveness) and also on growth
beyond the cyclical rebound, as gains from structural reforms are
projected to materialize gradually.
-
GREECE
14 INTERNATIONAL MONETARY FUND
Figure 1. Simulation Results, 201320
Source: IMF staff estimates.
-12
-8
-4
0
4
8
-12
-8
-4
0
4
8
2001 2004 2007 2010 2013 2016 2019
Real GDP Growth(Year-on-year percent change)
Program labor and product market reformsProgram labor market
reforms onlyBaseline (no reforms)
0
2
4
6
8
10
0
2
4
6
8
10
12
2013 2015 2017 2019 2021 2023 2025 2027 2029
Real GDP Growth (Cumulative deviation from reform baseline in
percent)
Program labor market reforms onlyProgram labor and product
market reforms
-3
0
3
6
-3
0
3
6
2001 2004 2007 2010 2013 2016 2019
Inflation (Percent)
Program labor and product market reformsProgram labor market
reforms onlyBaseline (no reforms)
-6
-4
-2
0
2
-6
-4
-2
0
2
2013 2015 2017 2019 2021 2023 2025 2027 2029
Inflation (Cumulative deviation from reform baseline in
percent)
Program labor market reforms onlyProgram labor and product
market reforms
-20
-16
-12
-8
-4
0
4
-20
-16
-12
-8
-4
0
4
2001 2004 2007 2010 2013 2016 2019
Current Account (Percent of GDP)
Program labor and product market reformsProgram labor market
reforms onlyBaseline (no reforms)
-3
0
3
6
-3
0
3
6
2013 2015 2017 2019 2021 2023 2025 2027 2029
Current Account (Deviation fromreform baseline in percent)
Program labor market reforms only
Program labor and product market reforms
-
GREECE
INTERNATIONAL MONETARY FUND 15
Figure 1. Simulation Results, 201320 (concluded)
-40
-30
-20
-10
0
10
20
30
-40
-30
-20
-10
0
10
20
30
2001 2004 2007 2010 2013 2016 2019
Investments(Year-on-year percent change)
Program labor and product market reformsProgram labor market
reforms onlyBaseline (no reforms)
0
5
10
15
20
0
5
10
15
20
2013 2015 2017 2019 2021 2023 2025 2027 2029
Investments (Cumulative deviation from reform baseline in
percent)
Program labor market reforms onlyProgram labor and product
market reforms
-30
-20
-10
0
10
20
-30
-20
-10
0
10
20
2001 2004 2007 2010 2013 2016 2019
Imports (Year-on-year percent change)
Program labor and product market reformsProgram labor market
reforms onlyBaseline (no reforms)
-2
-1
0
1
2
3
-2
-1
0
1
2
3
2013 2015 2017 2019 2021 2023 2025 2027 2029
Imports (Cumulative deviation from reform baseline in
percent)
Program labor market reforms only
Program labor and product market reforms
-30
-20
-10
0
10
20
-30
-20
-10
0
10
20
2001 2004 2007 2010 2013 2016 2019
Exports(Year-on-year percent change)
Program labor and product market reformsProgram labor market
reforms onlyProgram product market reforms only
0
2
4
6
8
10
0
2
4
6
8
10
2013 2015 2017 2019 2021 2023 2025 2027 2029
Exports (Deviation fromreform baseline in percent)
Program labor market reforms onlyProgram labor and product
market reforms
Source: IMF staff estimates.
-
GREECE
16 INTERNATIONAL MONETARY FUND
References Andersen, R., 2008, Modern Methods for Robust
Regression, Thousand Oaks, California: SAGE
Publications.
Barnes, S., R. Bouis, P. Briard, S. Dougherty, and M. Eris,
2011, The GDP Impact of Reform: A Simple Simulation Framework, OECD
Economics Department Working Papers 834, OECD Publishing.
Bernanke, B., M. Gertler, and S. Gilchrist, 1998. The Financial
Accelerator in a Quantitative Business Cycle Framework, NBER
Working Papers 6455, National Bureau of Economic Research, Inc.
Cacciatore, M., R. Duval, and G., Fiori, 2012, Short-Term Gain
or Pain? A DSGE Model-Based Analysis of the Short-Term Effects of
Structural Reforms in Labour and Product Market, OECD Economics
Department Working Papers, No. 948, OECD Publishing.
Christopoulou, R. and P. Vermeulen, 2012, Markups in the Euro
area and the US over the period 19812004: a comparison of 50
sectors, Empirical Economics, Springer, vol. 42(1), pp. 5377,
February.
Griffith, R. and R. Harisson, 2004, The link between product
market reform and macroeconomic performance, European Economy -
Economic Papers 209, Directorate General Economic and Monetary
Affairs (DG ECFIN), European Commission.
Gwartney, J., R. Lawson, and J. Hall, 2012, Economic Freedom
Dataset, published in Economic Freedom of the World: 2012 Annual
Report, Fraser Institute 2012
Hobza, A., and G. Mourre, 2010, Quantifying the potential
macroeconomic effects of the Europe 2020 strategy: stylized
scenarios, ECFIN Economic Papers 424, September.
Hoekman, B., H. L. Kee, and M. Olarreaga, 2001, Mark-ups, Entry
Regulation and Trade: Does Country Size Matter? CEPR Discussion
Papers 2853, C.E.P.R. Discussion Papers.
Kumhof, M., D. Muir, S. Mursula, and D. Laxton, 2010, The Global
Integrated Monetary and Fiscal Model (GIMF) Theoretical Structure,
IMF Working Paper.
Lee, J., G. Milesi-Ferretti, J. Ostry, A. Prati, and L. Ricci,
2008, Exchange Rate Assessments: CGER Methodologies, IMF Occasional
Paper No. 261.
Lusinyan, L. and D. Muir, 2013, Assessing the Macroeconomic
Impact of Structural Reforms the Case of Italy, IMF Working Paper
No. 13/22.
Milesi-Ferretti, G. and A. Razin, 1998, Current Account
Reversals and Currency Crises: Empirical Regularities, IMF Working
Papers, pp. 1-44, February.
-
GREECE
INTERNATIONAL MONETARY FUND 17
Novella, B. and M. Molnr, 2010, How large are competitive
pressures in services markets? Estimation of mark-ups for selected
OECD countries, OECD Journal: Economic Studies, OECD Publishing,
vol. 2010(1), pages 1-51.
OECD, 2012, OECD Economic Surveys: Euro Area, Paris.
Stewart, M. B, 1990, Union Wage Differentials, Product Market
Influences and the Division of Rents, Economic Journal, Royal
Economic Society, vol. 100(403), pages 1122-37, December.
Valencia, F. and L. Laeven, 2012, Systemic Banking Crises
Database: An Update, IMF Working Paper 12/163, International
Monetary Fund.
Zonzilos, 2010, Increasing competition in the Greek product and
labor markets: A macroeconomic assessment, Foundation for Economic
and Industrial Research.
-
GREECE
18 INTERNATIONAL MONETARY FUND
REVENUE ADMINISTRATION AND FISCAL CONSOLIDATION1
Greece has undertaken a very strong fiscal effort. During
200912, the primary balance improved by over 9 percent of GDP, or
15 percent in cyclically-adjusted terms. Further improvement of
some 4 percent of GDP is needed over 201316 in the
cyclically-adjusted primary balance to meet medium-term targets,
making this one of the largest fiscal adjustment programs to date.
This effort has been achieved primarily by raising tax rates to
high levels and reducing wages, pensions, and other spending.
Broadening the tax base and improving tax collection are also
essential pillars on which some progress has been made and on which
more is needed to ensure durability, fairness, and high quality of
the adjustment. This paper describes the problems, progress to
date, and agenda for work in Greeces revenue administration.
A. Why Improving Revenue Administration is Critical for
Greece
1. Further fiscal effort is needed to achieve debt
sustainability, and tax collection gains are expected to play a key
role. Going forward, to achieve the headline primary surplus target
of 4 percent of GDP by 2016, additional savings of 3 percent of GDP
will need to be identified. Almost half of this amount is slated to
come from better tax collection. Given already compressed
expenditure levels and high tax rates in Greece, the scope for
further measures in these areas is somewhat more limited.
2. Better tax collection is crucial to ensure broad-based public
support for the adjustment effort. The distribution of the revenue
burden across society, and perception that everyone is paying their
fair share, plays a key role in gaining support for the fiscal
adjustment mix. This is an issue at both the level of the
individual and at the level of businesses. To date in Greece, the
main adjustment burden has been carried by easy-to-tax salaried
employees and pensioners, while the richer, such as the
self-employed (e.g., doctors, lawyers), and other high wealth
individuals have continued to stay outside the tax net.
3. Stronger revenue administration would have macroeconomic
benefits. Weaknesses in the revenue administration have encouraged
small scale and inefficient economic units, which can evade taxes
more easily (often only the owner or the family is involved in
running the business). But such small scale operations tend to have
a very low return and productivity. Poor collection also implies a
need to maintain high tax rates in the formal economy, which hurts
the competitiveness of larger, export-oriented companies.
1 Prepared by Stephanie Eble (EUR) and Iva Petrova (FAD).
-
GREECE
INTERNATIONAL MONETARY FUND 19
4. The need for reform is urgent. The program assumed from its
onset that about 2 percent of GDP in savings would come from
revenue administration reforms. However, yields from those reforms
have been persistently shifted outward in time and reduced in
amount because of slow progress. Slow progress reflected a lack of
political commitment to take an unequivocal anti-evasion stance,
the absence of leadership in a very decentralized institution,
resistance to reform within the administration, weak implementation
of reform laws, excessive legal formalism, and nontransparent laws
and practices. Given the lead time in implementing those reforms,
without imminent action, the latest trajectory for revenue gains
could yet again be at risk.
5. The paper is organized as follows: The next section discusses
the scope for raising revenue and where in the collection process
is revenue being lost. The third section analyzes the rationale for
evading taxes in Greece. The fourth section discusses the reforms
underway to enhance revenue collection. The fifth section analyses
what factors have been important to bring those changes about,
drawing on country experiences. The final section concludes.
B. Revenue Losses in Greece and the Scope for Higher
Collection
6. Several empirical studies on the tax gap point to large scale
tax avoidance and evasion in Greece.
Schneider and Buehn (2012) in a cross-country study estimate the
size of the shadow economy during 19992010 at around 27 percent of
GDP, compared to an OECD average of 20.2 percent. Key explanatory
variables in the case of Greece are the degree of self-employment,
the indirect tax burden, the high unemployment rate, and tax
morale. Self-employment appears the most significantly correlated
variable with the size of the shadow economy in Greece, while tax
moralean important explanatory variable in other advanced
countriesdoes not score as high as one might desire.
Based on banks perception of true income, Artavanis et al.
(2012) estimate that annual unreported income exceeds 28 billion.
They find that occupations most prone to tax evasion in Greece are
lawyers, doctors, accountants, private tutors, and engineers.
Further, Papageorgiou et al. (2011) find that, despite high
statutory rates, effective tax rates in Greece are much lower than
in the euro area, suggesting a high level of tax evasion and
avoidance.
-
GREECE
20 INTERNATIONAL MONETARY FUND
7. There are many (often discretionary) exemptions in the tax
code that erode the tax base. While there has been a significant
effort at streamlining non-productive and distortive tax
exemptions, important ones in the areas of personal income, VAT and
capital taxation remain, amounting to 2 percent of GDP in 2011. In
particular, for VAT, a super-reduced rate of 6.5 percent applies to
hotels, medicine and books/news papers. For all Aegean islands, the
VAT rates are reduced by 30 percent, i.e., a VAT of 16/9/5 percent
compared to 23/13/6.5 percent on the mainland. With respect to PIT,
remaining exemptions relate to having the status of a pensioner and
disability, which have been subject to widespread misuse (the
ongoing review of the disability status is only expected to be
completed by 2016). CIT exemptions are subject to a high element of
discretion and generous rules (e.g., treating government grants as
expenditure for tax purposes, offsetting of income from other
investments, etc.) and are not incorporated in the medium-term
fiscal framework. Overall, there is scope for further streamlining
non-productive, untargeted, and discretionary tax expenditures, and
possibly replacing them with accelerated depreciation rules.
8. Further, tax avoidance schemes are used to reduce tax
obligations. Ambiguities, loopholes, and interpretations of the law
are frequently exploited to reduce tax obligations. For example,
social security contributions for the self-employed are much lower
than those for salaried employees, since the contribution to their
funds is not related to income but to years of participation. This
makes it attractive to work as self-employed, including because of
the weak link between benefits and contributions. Also, since
imputed income is often used for the purpose of taxation, there is
less of an incentive to declare ones true income, especially as
certain real estate investments and equity capital contributions
are exempted from the calculation of the presumptive tax base.
Furthermore, real estate transaction taxes use the objective,
rather than the market, value as the tax base. The discounted VAT
rates for the Aegean islands are also open to tax avoidance
opportunities, as purchases consumed on the mainland can be
channeled through them. Finally, transfer pricing is commonly used
by Greek companies to route profits to low-tax jurisdictions.
2008 2009 2010 2011
1. Direct taxes 5,758 4,653 3,501 3,175 Income taxes 3,814 2,875
2,068 1,742
PIT 3,522 2,642 1,964 1,625 CIT 292 233 104 116
Capital (incl. property) 1,944 1,778 1,433 1,433 2. Indirect
taxes 3,086 3,551 3,327 2,354
Transaction taxes 682 603 491 1,096 VAT 565 543 455 1,077 Other
(vehicles) 118 60 36 19
Consumption taxes 2,404 2,948 2,837 1,258 Total (1+2) 8,844
8,204 6,829 5,529
Sources: Ministry of Finance; and IMF staff estimates.
Tax Expenditure(Millions of euros)
-
GREECE
INTERNATIONAL MONETARY FUND 21
Table 1. Where is Greece Losing All Its Tax Revenue?
9. Standard indicators suggest that there remains significant
scope for raising revenue in Greece via stronger collections:
VAT and CIT productivity in Greece are very low by EU standards,
in part reflecting reduced and multiple rates and exemptions.
Channel Explanation Estimated Annual Loss
Tax exemptions Several exemptions for VAT and PIT.Discretionary
investment incentives.
2.5% of GDP in 2011
Tax avoidance schemes Status of self-employed versus salaried
employee.Presumptive taxation and objective values.Location
specific VAT rates.
.
Non-declaration of taxes Non-declared labor income and income of
liberal professions.Non-filling of VAT.Unrecorded and under
recorded transactions.
4%5% of GDP (Avanidis et al)
Assessed debt through audits is discounted
No audits of the auditors.No tracking of the originally detected
amount, discounts, and finally assessed amount.Up to 80 percent
reduction of interest and penalty charges on undeclared
liabilities.
0.1% of GDP (only a small share of tax liability is collected
through audits)
Large share of assessed debt remains unpaid
Insufficient focus of collection effort on new and collectable
debts. 5% of GDP
Administrative delays Extension of filing/payment deadlines.Many
and generous deferred payment arrangements.Lengthy court
proceedings.
..
Source: IMF staff estimates.
-
GREECE
22 INTERNATIONAL MONETARY FUND
ProfessionIncome
Multiplier
Doctors and medicine 2.45Engineering and science 2.4Education
2.55Accounting and financial services 2.22Law 2.24Fabrication
2.26Media and art 2.22Lodging and restaurants 1.99Construction
1.85Business services 1.62Transport 1.51Agriculture 1.75Personal
services and pharmacy 1.49Retail 1.27Others 1.22
Source: Artavanis, et al., 2012.
Implied Income to Reported Income
Tax debt in Greece is very high by international standards, as
tax payers fail to pay even what has been assessed.
A large share of income is never declared. Inspections by the
Labor Inspectorate (SEPE)
suggest that about every third employee is not registered.
Furthermore, the average declared income of the self-employed is
close to the minimum income and well below the GDP per capita,
suggesting that a large share of the actual income remains
undeclared. Sectors with high propensity to use undeclared work are
tourism, catering, construction, agriculture, homecare, and
commerce. According to Artavanis et al. (2012), actual income could
be up to 2 times larger than that declared by doctors, engineers,
and private tutors.
-
GREECE
INTERNATIONAL MONETARY FUND 23
10. Because of weak enforcement, a large share of the assessed
tax remains unpaid. In 2012, new debt of 13.1 billion accrued, or
25 percent of total assessed taxes. Of this, only some 1.4 billion
were collected (less than 12 percent of revenue). Similarly, for
social security contributions, the gap between actual collections
and the assessed amount reached 15 percent for IKA, 25 percent for
the self-employed fund ETAA, and 65 percent for the agricultural
fund OGA in 2012. Overall, unpaid obligations reached 56 billion
for taxes and more than 12 billion for social security
contributions at end-2012. Enforced collection focuses on the
oldest claims, irrespective of their low collectability compared
with new debts. Further, the available instruments to enforce
collection are inadequate by international standards. Since 14
percent of the debts represent 92 percent of debt value, there is a
need for scarce human resources to focus on these debts and for
other debt to be subject to automated interventions. As
insufficient resources are dedicated to the collection of new debt,
such debt quickly becomes uncollectable. Technical assistance
advice to dedicate greater resourcesat least 10 percent in local
tax officesto collection has also remained unaddressed, and the
workforce continues to be dedicated mostly to the assessment of
debt.
C. What Explains the Large Amount of Tax Evasion in Greece?
11. Evasion has generally been modeled as driven by costs and
benefits. On the benefit side are the lower tax payments explained
by higher tax rate; on the cost side are penalties combined with
the probability of actually being detected. The optimal tax evasion
behavior can be described by the following equation (Yitzhaki,
1974):
1
where, YA denotes audited income, YU denotes unaudited income, p
represents the probability of detection, t is tax rate, and is the
penalty on evaded income.
12. In Greece, the benefits of evasion are comparatively high
and the costs are comparatively low.
High tax rates. High rates create a strong incentive to not
declare, since the marginal benefit from hiding any income is high.
Tax rates in Greece are at the top end of the OECD range. A tax
wedge on salaried employees of 43 percent compares with an OECD
average of 26 percent, and the standard VAT rate of 23 percent
places Greece among the three advanced countries with the highest
VAT rates.
-
GREECE
24 INTERNATIONAL MONETARY FUND
Low probability of detection. Historically, there have been a
number of problems with conducting audits, such as: (i) focusing on
non-filers and non-active companies, with minimal prospects of
collecting additional assessed taxes, leaving high-risk cases
unlikely to be audited; and (ii) limited probability of detecting
undeclared tax liability when an audit is carried out, due to
limited use of third party information to detect inconsistencies
between declared and actual wealth and income, and lack of access
to bank account information. Audits focus on bookkeeping
formalities rather than assessment of tax liabilities, reducing
auditor productivity.
Low effective penalties. While penalties stipulated in the law
are generally high in Greece, the probability is very low that
these penalties will be fully applied and enforced. For instance,
when an audit order has already been sent out, one can still apply
for an amnesty on audits. In addition, if an audit detects
undeclared liability, the local tax office manager may reduce the
penalties for non-compliance (about 100 percent of the tax
liability) and the interest for late payment (standard rate per
month of 1 percent, not compounded) by up to 80 percent. If the
taxpayer settles the final liability immediately, the taxpayer
receives a 5 percent cash rebate on the total amount paid and faces
little risk of prosecution. If this liability becomes overdue the
taxpayer can participate in an installment scheme, which offers
additional discounts to the penalties or interest by 50 percent if
paid immediately. This reduces the effective interest rate well
below commercial lending rates, making it cheaper to not declare
taxable incomes even in the case that the actual tax liability is
eventually detected (Table 2, Figure 1).
-
GREECE
INTERNATIONAL MONETARY FUND 25
Not Declared
0 Euros Due
Very low propability of being audited
Application for Audit Amnesty
Low probability of detection of
non-declared tax liability
Interest (1 percent per month) and penality are
reduced by up to 80 percent
Immediate Payment Incentive: 5 Percent
Discount
With Installment Scheme: up to 36 monthly payments
When Litigated: 50 Percent Due Immediately and 10
Year Extension on the RestApply for Amnesty
Declared
Immediate payment: Cash discount of 1.5%
Deferred payments: 3-7 monthly payments
(interest free)
If No Payment
1 percent Interest accrued per month
Non-payment: Prison for Tax Debt Above 5,000 Euros
Payment: Installment Scheme
Table 2. Costs of Alternative Tax Payment Schemes: Bank Loan
Deferral Scheme or Payment Following Audits
Figure 1. Tax EvasionsA Decision Tree
Year 1 2 3 4 5 6 7 8 9 10
1. Tax liability declared and financed through unsecured bank
loan (12 percent interest)
100 112 125 140 157 176 197 221 248 277
2. Tax liability declared and financed through installment
scheme (1 percent interest per month)
100 112 124 136 148 160 172 184 196 208
3. Non-declared tax liability detected Tax liability 100 100 100
100 100 100 100 100 100 100
Penalty (about 100 percent of tax liability) 100 100 100 100 100
100 100 100 100 100Standard interest rate (1 percent per month, up
to 200 percent of tax liability)
24 48 72 96 120 144 168 192 200 200
Total liability (tax liability plus surchages) 224 248 272 296
320 344 368 392 400 40066 percent discount on penalty and interest
charges (up to 80 percent)
-83 -99 -115 -131 -147 -163 -178 -194 -200 -200
Final liability due 141 149 157 165 173 181 190 198 200
200Immediate payment (5 percent discount on total amount) 134 142
150 157 165 172 180 188 190 190
If only 50 percent of actual tax liability detected 67 71 75 79
82 86 90 94 95 95Payment of overdue obligations under installment
scheme (50 percent discount on interest and penalties)
121 125 129 133 137 141 145 149 150 150
If only 50 percent of actual tax liability detected 60 62 64 66
68 70 72 74 75 75
Source: IMF staff estimates.
(Example for actual tax liability of 100 euros)
Payments Due
-
GREECE
26 INTERNATIONAL MONETARY FUND
Amnesty. Frequent use of amnesty schemes has created
self-fulfilling expectations of more generous future schemes. For
instance, under the last big scheme of 2010 a minor additional tax
liability could be paid in exchange for not having the tax returns
audited.
13. Other social and economic factors identified in empirical
studies can also help to explain evasion in Greece.
Weak social norm of compliance and a tax system perceived as
unfair. For instance, Andreoni et al. (1998) find that the
taxpayers perception of the fairness of the tax system and burden,
and that there are moral rules and sentiments are key factors to
explain compliance. With respect to Greece, the lack of political
will to tackle tax evasion of certain groups is often perceived as
an element of injustice of the system.
High dissatisfaction with government services and public goods.
If the individual does not perceive that he gets anything in return
from the government, the incentives not to pay taxes are high.
Greece has an oversized and inefficient public sector. The level of
corruption in the government is high, with Transparency
International ranking Greece as the most corrupt country in the EU.
The quality of the services delivered by the administration to its
citizen is low, requiring often side-payments, queuing time and
excessive bureaucratic procedures. As a consequence, citizens
depend heavily on costly private services, including for education,
health and motorways.
Pressure for business survival. Companies face competition from
a large share of companies that do not charge or pocket the VAT, do
not pay the high labor tax wedge of 43 percent and avoid CIT/PIT on
their income, and are thus able to provide the product or service
at much lower prices. Further, the cyclical impact of the deep
recession on companies is putting further pressure on companies to
evade taxes in order to survive.
-
GREECE
INTERNATIONAL MONETARY FUND 27
Table 3. Irrationality of Paying Taxes
14. The result of the costs and benefits in Greece is widespread
evasion. Over time, economic organizations have shifted to
facilitate to this. The shift in economic organizations has in turn
fed back onto the tax administrations ability to detect and
enforce, creating a vicious cycle.
15. Indeed, in Greece, about 35 percent of the labor force is
self employed, compared to a European average of 14 percent. This
is mainly driven by tax avoidance and evasion schemes, as self
employed carry a lower tax burden and find it easier to hide their
income. Further, in Greece about 55 percent of the employees work
in companies of less than 10 employees.
Compliant Tax Payer Tax Avoider and Evader
Staying in business
Charges a 23 percent VAT rate on the sale of goods and services,
insures all its employees with a tax wedge of 43 percent and
declares all its profits. The company will be subject to unfair
competition from businesses not being compliant with their tax
obligations. Eventually it will go out of business.
Does not impose any VAT, or pockets the VAT, and employment is
informal. Even though the company is not very productive, it can
stay in business because it benefits from very high margins.
Paying penalties for undeclared work is cheaper than paying
contributions
Pays social security contribution for all its employees. The tax
wedge on labor amounts up to 43 percent.
On average, every company is visited by the labor inspectorate
every ten years. If undeclared employment is detected, the average
penalty for an informal employee is Euro500. This is only one month
of payment in social security contributions and PIT for a minimum
income earner. By law, the labor inspectorate can close business,
but this is rarely done.
Not declaring has lower penalties than declaring and not being
able to pay
Declares all tax obligations. If he/she does not have the
resources to pay the tax debts, borrows from the bank or accrues a
tax liability with a monthly interest rate of 1 percent, and the
risk of prison charges.
It is estimated that only a small part of the undeclared income
is detected in tax audits. The undeclared tax liability as a result
of the tax audit assessment will be due only in up to 36 monthly
installments. If the taxpayer decides to pay immediately he will
receive a cash rebate of 5 percent. The interest and penalty
charges can be negotiated to a minimum and if the taxpayer actually
pays the tax due there is rarely any risk for him to be
prosecuted.
Amnesty schemes are offered
Stays current on all tax obligations.Waits for the next amnesty
scheme. If he/she cannot comply with the payment schedule of the
amnesty scheme, then drops out and waits for the next one.
Installment schemesPays tax obligations in time, and if needed
borrows at market rates from the bank to meet the tax
obligations.
Takes advantage of very generous installement schemes, which do
not compound interests and offer generous discounts if the total
tax liability is repaid.
Going to court is better than paying taxes
After the additional tax debt has been assessed through an audit
this must be paid within 36 monthly installments.
Tax payer goes to Court with a projected hearing data in--on
average--10 years. In a large share of the tax cases, the court
decides in favor of the tax payer.
Deadlines are always extended
Files and pays taxes in time. Waits for an extension of the
filing and payment deadlines or an installment scheme.
Source: IMF staff.
-
GREECE
28 INTERNATIONAL MONETARY FUND
D. The Way Forward
16. In general, the organization, procedures, and operations in
every tax administration change over time. Reforms are needed to
keep up with new challenges in tax compliance and with evolving
needs for taxpayer services. Revenue administration reforms have
also picked up in the last several years in many advanced countries
with a view to reduce costs and support fiscal consolidation. A
reform strategy depends on the size of the tax gap and the already
existing organizational effectiveness of each revenue
administration (Table 4, Silvani and Baer, 1997).
17. Greece needs a holistic approach to revenue administration
reform to achieve effectiveness and flexibility and respond to the
outstanding fiscal consolidation needs. A comprehensive reform
strategy is the only possible course, given deep-rooted problems of
mismanagement, staff qualification and integrity, and lack of
taxpayers compliance. It is also the only solution to aligning
Greece with advanced country practices. Such a comprehensive
strategy was provided in 2011 technical assistance advice (IMF,
2011), aiming at: (i) reducing tax evasion and the shadow economy;
(ii) improving effectiveness in revenue collection; (iii)
eradicating corruption to restore the tax authorities credibility
and taxpayers confidence; (iv) encouraging voluntary tax
compliance; and (v) improving the redistributive effects of the tax
system. As problems of political interference and resistance also
appeared to run deep, in 2012, technical assistance further
determined the need for a minimum level of autonomy of the revenue
administration to ensure that a reform strategy can be carried
forward (IMF-EC, 2012).
-
GREECE
INTERNATIONAL MONETARY FUND 29
Table 4. Criteria for Tax Administration Reform
Tax Gap
Tax Administration Features Country Examples
Reform Type
30% Lack of financial and material resources Poorly qualified
and trained staff Ineffective procedures Non-implementation of
measures to
improve compliance High turnover of technical staff Corrupt
practices No taxpayer services
Peru, 1991 Bulgaria,
2000 Georgia, 2005 Greece, 2010
A radical and comprehensive reform to modernize all laws ,
systems, and processes affecting tax collection and creating a well
equipped and trained workforce to efficiently and transparently
administer the tax system.
Sources: Brondolo et al., 2008; Durand, 1996; EC, 2009; IOTA Tax
Tribune, Issue 28, 2011; OECD, 2011; Silvani and Baer, 1997; and.
World Bank, 2000 and 2009.
18. In light of the broader problems with the tax system in
Greece, the comprehensive reform strategy extends beyond tax
administration reforms. To improve the effectiveness of the tax
administration, it seeks to: (i) simplify the existing tax
legislation; (ii) simplify tax procedures; and (iii) reform the
governance, organization, and operations of the tax administration
(Figure 2). The simpler tax policy regime and tax procedures rules
provide greater incentives for taxpayers to comply with their tax
obligations and are easier to administer.
-
GREECE
30 INTERNATIONAL MONETARY FUND
Figure 2. Greece: Revenue Administration Reform Strategy
Simplifying the tax policy regime. A new income tax legislation
adopted in January 2013
reduced the number of personal income tax brackets from 8 to 3
and eliminated a number of tax credits and deductions. It also
reduced the number of tax rates at which different types of capital
income are charged. Reforms in the coming months will overhaul the
existing Income Tax Code and its amendments into a single piece of
legislation and eventually consolidate provisions on tax incentives
currently existing in other laws, and simplify the progressive
surcharges on rental taxation and overall income. By June 2013,
property taxes will also undergo a major reform consolidating the
existing state-level property tax (FAP), the extraordinary real
estate tax collected by PPC through electricity bills, and the real
estate transaction tax. This reform will simplify the multiple
rules currently used to assess property tax obligations, establish
a common broad base for property taxation, and eliminate widespread
uncertainties in tax assessments. Going forward, a simplification
of the VAT regime is also needed to reduce the opportunity for
arbitrage toward lower rates and create a more stable tax base over
the business cycle.
Streamlining tax procedures. The new tax procedures code (TPC)
takes over legal provisions that previously existed in the income
tax code, tightening provisions for audit, assessment, provision of
information, collection enforcement, and dispute resolution. These
provisions are intended to clarify filing and payment procedures
and set clear deadlines, ease the collection of information
relevant for tax assessments, strengthen the ability of the revenue
administration to place a lien on a taxpayers property and dispose
of the property to satisfy a tax obligation, impose an interest
that compensates the government for the opportunity cost of belated
tax
Simplify the Tax Regime
Simpler tax systems are easier to administer and result in
higher compliance levels.
Strengthen the Revenue Administration
A strong and autonomous, but accountable, management and
qualified and motivated staff provide a backbone for high
performance. A
function-based organization is more efficient. Effective
auditing and
collection discourage tax evasion.
Streamline and Clarify Tax Procedures
Good tax procedures discourage delay in tax payments and provide
a
system of sanctions and penalties that encourages quick
settlement of
arrears.
-
GREECE
INTERNATIONAL MONETARY FUND 31
revenue, and charge penalties on late tax and filing obligations
in a transparent manner. The bookkeeping code was also simplified
in November 2012, and is expected to be streamlined further shortly
after the adoption of the TPC.
19. Strengthening revenue administration has several
components:
Operational reforms. There is a need to improve detectability of
tax evasion by: (i) adopting segmentation strategies that focus
resources where the tax risks are the greatest (e.g., largest
taxpayers, high-wealth individuals, and high income self-employed);
(ii) improving the incentives and qualifications of staff in these
units; (iii) increasing collection of third party information; and
(iv) using risk-based approach to audit. Enforcement also needs to
be strengthened by: (i) improving debt collection; (ii) adhering
strictly to the no amnesty commitment; (iii) introducing
installment arrangements in line with international practice; (iv)
adhering to filing deadlines; (v) improving dispute resolution and
judicial process; and (vi) developing taxpayer assistance
strategies to improve voluntary tax compliance. Legislative changes
to upgrade operations have already been made. Legal constraints to
write-off uncollectable amounts of tax debt have now been removed.
Legal obstacles to auditing those cases that represent the highest
revenue risk have also been dismantled. Legislation creating formal
audit centers for large taxpayers and high-wealth individuals has
been passed. Legislation to use indirect audit methods and
establish the actual income of taxpayersrather than rely
excessively on presumptive methods of taxationhas also been
adopted. Finally, a new legislative framework for installment
schemes has been adopted, which is intended to depart from the
vicious cycle of extending amnesty schemes.
Autonomy and organization. There is a need to remove completely
political interference from revenue administration. A law passed in
2012 established the General Secretariat for Public Revenue and
required transferring powers from the Minister to the Secretary
General over implementation of customs and tax laws related to
revenue collection; provided for the Secretary General to appoint
heads of units; and required that the Secretary General be a
non-political appointment for 5 years. Further legislative changes
in March and April 2013 gave powers to the Secretary General to
make changes of organizational units without constraints imposed by
the Ministry of Administrative Reform (MAREG). Accountability of
the revenue administration is strengthened by the provision to
establish an advisory board to the Secretary General.
Resources and staffing. The Greek revenue administration is
agingwith more than 50 percent of staff exceeding 50 years of age.
Skills are also lacking across main functions, such as audit and
debt collection. However, upgrading staff is notoriously difficult
due to a low-entry level pay and a flat grade for auditors, which
do not create sufficient incentives for existing and new staff to
perform complex tasks. For example, in comparison with revenue
administrations in other advanced countrieswhere the minimum
remuneration of an auditor is about 134 percent of GDP per capita,
tax auditors in Greece are paid significantly lessat less than 50
percent of GDP per capita. Auditor pay scale distinguishes multiple
grades, and auditor remuneration reaches 340 percent of per capita
GDP at the highest grades (OECD, 2006). Under newly passed
-
GREECE
32 INTERNATIONAL MONETARY FUND
legislation, the Secretary General is now allowed to determine
the grading and promotion system of the revenue administration,
which will allow changing staff incentives and encouraging skilled
staff to take challenging audit and debt collection functions. The
appointment process for staffwhile is still under the control of
ASEP and constrained by the overall general government hiring
constraintswill also be determined by the Secretary General. A code
of ethics for the revenue administration that determines standards
for professional integrity of staff and an anti-corruption plan for
the entire public administration have also been adopted.
20. Despite significant legislative effort, the reform continues
to be challenging. While started in 2010, it has continuously faced
both political and internal resistance to change. The reform has
been marred by lack of direction, management, and control from
headquarters, especially in periods of political uncertainty: the
revenue administration had no head for most of 2012. The process to
reshape the organization by scaling down powerful local tax offices
and toward a function-based organization has been very slow. As a
result, the revenue administration continues to operate under
mostly decentralized control, in which eradicating corruption
remains a major challenge. Efforts to upgrade the quality of audit
staff through both external recruitment and internal certification
have mostly failed due to union resistance and management inability
to effectively reassign staff.
E. Change: How and How Quickly Can It Happen?
21. Other countries offer useful lessons on the key ingredients
for successful transformations:
Unequivocal political support to reform is a common factor for
success (Table 5). The reform undertaken in Hungary in the early
1990s, which sought to improve collection and audit capacity and
upgrade the IT system, saw rapid initial progress under the
powerful supervision of the president of the tax administration
agency. The agency had been established in the late 1980s as a
semi-autonomous organization, with non-political leadership having
sufficient powers to implement organizational reforms. However, a
government change led to a change in governance, as the position of
the president of the tax administration became a political
appointment. Political interference in agency affairs continued
throughout the mid-1990s, which caused reform delays and eventually
hampered the adoption of new audit selection methods.
-
GREECE
INTERNATIONAL MONETARY FUND 33
Table 5. Country Experiences with Institutional Reforms
Country Type of Reform Period
Revenue Administration Reforms
Latvia Consolidating regional offices and reducing the structure
of the organization from 23 to 15, mostly functional, units.
200809
Bulgaria
New unified revenue administration; consolidating 340 offices
into 29; upgrading functions; merging SSC collection into tax
administration; IT platform.
Further reduction of offices from 29 to 7.
200009
2009
Chile Activating an anti-evasion plan, expanding large taxpayer
audit staff; new decentralized work force for cash economy
measures; information technology upgrades.
200209
Georgia
Reorganized tax headquarters and field units; New tax code and
simplified administrative procedures; abolishing compulsory audit
of all annual filings and adopting a risk-based approach to tax
audits; compulsory electronic filing; new two stage administrative
dispute resolution mechanism.
200507
FYR, Macedonia
New tax procedures law; restructuring of headquarters and field
offices; Creating large taxpayer office; flexibility in hiring,
firing, rewarding staff; integrating collection of social insurance
contributions to tax agency.
200509
Peru Creating an independent revenue administration agency
directly reporting to the president with full budget
flexibility.
199195
Hungary
Comprehensive: reorganizing the HQ office; creating an LTU;
improving VAT and PIT procedures: new arrears management
approach.
IT upgrade; Improving audit selection.
199396
199698
Sources: Durand, 1996; IOTA Tax Tribune, Issue 28, 2011; and
World Bank, 2000 and 2009.
A reform program office charged with reform implementation is
needed to break through
stiff resistance. Many modernization programs fail due to
failure to manage the complexity of reform while keeping up with
the overwhelming nature of the day to day business demands on a few
key leaders. The reform of the Peruvian revenue administration in
1991 is one of the best examples of efficient transformation within
a public service. The country was engulfed in a crisis that
exacerbated already deeply bureaucratic and corrupt practices. The
policy team in charge of the revenue administration reforms had the
highest political support and comprised a number of reformed-minded
members with spotless reputation and strong technical skills. The
team developed and carried out a sweeping plan since experience had
already shown that incremental changes were prone to setbacks,
resistance, loss of political support, and reform fatigue. The
element of surpriseboth for taxpayers who saw unyielding new
revenue
-
GREECE
34 INTERNATIONAL MONETARY FUND
administration staff and for the labor union that resisted the
changeyielded results and was possible due to a careful selection
of a new generation of university educated recruits to handle
difficult compliance cases.
Timing is essential. Swift reform implementation has been shown
to secure an early gain from demonstrating political will and value
for money. Bulgarias transformation of the revenue administration
took almost 10 years. Despite impressive efficiency gains made
during the period, the revenue agency remained hampered by large
operating costs in the wake of the global financial crisis. Of 29
existing regional offices, only 7 collected 80 percent of revenues.
A decision to further rationalize the organization of the revenue
agency would support the governments objective to keep public
administration costs low while improving efficiency. However, as
memories of the previous long-lasting reforms were still fresh, the
reorganization had to be completed rapidly to avoid resistance,
high staff turnover, drop in morale, and efficiency loss. This was
done in four months at the end of 2009, containing reform costs
within the current fiscal year and budget envelope.
Reforms should focus on the areas that have the highest revenue
potential. Reforms in a number of countries have shown large
revenue gains once the institutional organization was shifted from
a regional to a function based model and refocused on large tax
payers, debtors with large tax liabilities to the government, and
stronger enforcement. In Indonesia, given the lack of capacity to
introduce comprehensive reforms, the authorities initially focused
only on collection and audit of large taxpayers. Strengthening
enforced collection raised the revenue ratio by percent of GDP,
while the establishment of a large taxpayer unit increased revenues
by percent of GDP over 4 years. Revenue gains could be higher over
a longer period of timeBulgaria increased tax and SSC collection by
5 percentage points of GDP, and VAT collection alone, where
non-compliance had been rampant, by 3 percentage points of GDP over
a 6-year period. In the most successful cases, such as Georgia
where tax revenues increased by 6 percent of GDP in 2005-07,
substantial revenue gains could be achieved even within a
medium-term horizon. Experience in the European Union more broadly
has also shown that rationalization of revenue administration costs
has been associated with an increase in the revenue ratio in the
subsequent year, most of which is in indirect taxes.2 Therefore,
there is a strong argument to focus resources on narrowing the VAT
gap.
2 Based on panel data estimation using annual data for the 27 EU
countries during 2007-2011 with the dependent variable being tax
revenue to GDP and independent variables being cost of tax
administration as a share of GDP and the number of hours needed to
complete tax filing and payment requirements. The results show that
a decline in tax
(continued)
-
GREECE
INTERNATIONAL MONETARY FUND 35
Staff reductions in the revenue administration can help improve
the quality of staff and overall efficiency. Cost reductions have
been a main factor to undertake revenue administration reforms in
advanced countries and in several emerging economies in recent
years (OECD 2011). In 2008, the Latvian revenue administration
refocused the organizational model on functional basis, removing
the need to maintain large support staff in regional offices. The
reorganization led to a staff reduction by 10 percent while the
reform was still ongoing in 20082009, and an additional reduction
of 13 percent in the following 2 years. The rationalization of
regional offices in Bulgaria brought a reduction of staff by 8
percent in 2009 alone, which added to a 13 percent reduction from
the previously completed reforms. In some cases, the
rationalization also required demotion of mid-level managers due to
the smaller office network. The reforms in Peru in the early 1990s
led to the dismissal of 2/3 of the existing staff. In all cases,
resources released from the reduction of staff were used to improve
the incentives, qualifications, and career prospects of employees
who were reallocated to the newly created functional units.
A new generation of tax administration officials changes the
culture and creates a new public image of the revenue
administration. During the reform carried out in Peru in the early
1990s, strict criteria were set for recruitment of new staff. More
than 15,000 applicants were tested, many interviewed, and only 3
percent selected. The recruitment was carried out by the revenue
administrations training institute, which then completed
specialized training in legal matters, audit, tax collection, and
ethics. Comprehensive reforms in other countriesGeorgia, Bulgaria,
and the Baltic countriesalso introduced new hiring, grading,
promotion and training policies seeking to establish a new
professional workforce. Codes of ethics, strict reprimands for
accepting giftseven outright dismissal as in Peruand regular
rotation of auditors also helped raise the level of professionalism
and integrity in these countries.
Passing through the organizational reforms allows the revenue
administration strategy to focus on better compliance even in the
most difficult circumstances. The 2009 crisis found Latvia with a
23-percent reduction in revenues and almost 3 percentage point of
GDP reduction of the revenue ratio relative to 2008. Nonetheless,
the 20082009 reforms allowed taxpayer categories with large risk of
non-compliance to be identified more easily. It also helped
reallocate resources faster to tasks that would require greater
emphasisreinforcing tax debt recovery, filing processes, audit, and
measures to combat criminal activities. Before the crisis hit,
Estonia had also undergone organizational changes merging core
functions, reducing support staff, and focusing on efficient tax
debt collection and a new tax refund system. These changes provided
a good platform to handle large tax arrears, including by providing
installment plans to viable enterprises. Small debt installment
arrangements were provided on the basis of automatic risk criteria,
while specialized staff was reallocated to handle more complex
cases, where assessment of business viability was more difficult.
The new VAT refund system also allowed
administration costs by 0.1 percent of GDP is associated with an
increase in the revenue ratio by 0.8 percent of GDP and in indirect
taxes by 0.7 percent of GDP in the subsequent year.
-
GREECE
36 INTERNATIONAL MONETARY FUND
more efficient scrutiny of high-risk VAT refund claims. A plan
to improve compliance in Peru was also possible only after the
first stage of reorganization was completed. This is because a good
compliance plan relies both on strict law enforcement and capacity
of revenue administration to provide quality taxpayer services.
Tax rate reductions should be used to support the reforms. The
improvement in voluntary tax compliance and increased collection
efficiency in Bulgaria allowed reduction of tax and social
contribution rates. This helped reduce business informality by
about 30 percent in 20022008. Reduction in tax compliance costsa
comprehensive tax reform combined with introduction of e-filinghas
helped buoy tax revenues in Georgia and reduce the tax burden on
companies by more than 20 percentage points of company profits
since 2007. In advanced economies, experience in the EU since 2007
has also shown that reforms that have aimed at reducing revenue
administration costs and compliance costs for the taxpayersin terms
of hours spenthave lead to an increase in revenue and a subsequent
reduction in the tax burden on companies. In particular, IMF staff
estimates suggest that a decline in the hours needed for compliance
by 100 could increase the revenue ratio by about 0.2 percent of GDP
and allow a reduction in the tax burden on businessesthe ratio of
total taxes payable by a business as a share of profitsby 1.3
percentage points.3
F. Conclusions
22. Experience in other countries shows that a change in tax
administration requires significant effort and could take time. The
review of the empirical factors has shown that the prevalence of
tax evasion in Greece is partly the result of administrative
capacity and partly because of fundamental flaws in the design of
the incentive and penalty structure of tax administration. While
several factors are less under the control of the Greek government
(e.g., decentralized economic structure), others can be addressed
immediately through changing and enforcing laws:
Simplifying the tax system to remove tax avoidance schemes and
ambiguities. This provides certainty for the taxpayers and reduces
the administrative costs for tax administration, including
litigation costs for tax assessment disputes.
3 The results are based on a panel data estimation using annual
data for the 27 EU countries during 2007-2011 with the dependent
variable total tax rate and the independent variables cost of tax
administration as a share of GDP and the number of hours needed to
complete tax filing and payment requirements.
-
GREECE
INTERNATIONAL MONETARY FUND 37
Establishing a legal and institutional framework that makes it
rational to comply with the tax law. This includes increasing the
effective penalties, improving the enforcement mechanism, and
adhering to a policy of no new amnesty schemes. O